The present disclosure relates to image processing and analysis, particularly the assessment of the texture of skin, such as the facial skin of humans.
In selecting medical or cosmetic treatments or products to be applied to human skin, it is helpful to assess the condition of the skin, before, during and after such applications. A condition of the skin that is often of interest is the texture of the skin, which is typically assessed by human observation. Whether made by the subject or even a professional, such an observation-based assessment, however, is largely subjective, difficult to quantify, and prone to wide variability. Additionally, some time and effort is required on the part of the observer to make, communicate, and record their assessment. The aforementioned issues are particularly problematic, for example, in studies where products or treatments are applied to large numbers of subjects and the efficacy thereof is to be assessed.
Facial skin texture is a combination of features (pores, bumps, etc.) that make up the overall surface topography in non-wrinkled areas of the face, like areas of the cheek, for example. Texture does not include color components, pigment information, underlying curvature or structure, and lighting gradients. Such features further complicate the objective assessment of skin texture.